arXiv:1806.10385v2 [math.GN] 21 Sep 2018 htfrmtial pcshspoet seuvln ooeconside one to equivalent is Each property [19]. his in Menger spaces pr Hurewicz metrizable because for used is that property) Menger (the name current ucvr.TeMne rpryi h eks n mn h so the among one weakest the is principles c property a selection Menger has cover The open every subcover). is, lindel¨ofness (that implies latter the and ler 3] obntrc fdsrt usae 2,adTkyrela Tukey and spaces [2], [14]. Menger the subspaces compacts Ramsey of discrete [12], topic. hyperspaces of between forcing the combinatorics as to areas [34], such introductions algebra in detailed applications found for recently 32] 26, 25, X hr r w omlspaces we nega pre normal answer precisely, two the is More spaces are property topological there Menger general [24]. For the in whether products. finite question initiated the principles on selection concentrate and fusion with hr xssasequence a exists there egrpoet i at hyhv h togs combinatoria strongest product whose the far), have thus they considered fact, property exist (in set-theoret there ambient property CH concent the under Menger we 3.2] to Theorem which sensitive [22, be on by to Indeed, case, out This turned sequel, spaces. the metrizable to attention our e [29, see hnteMne n uhthat such one Menger the than 03E05. n 34N5 o eeosspotfrti research. this for support generous for 2374-N35) I and egrsae ffreeysequence every for if space) Menger a 2010 e od n phrases. and words Key = h uhrwudlk otakteAsra cec udFF(Grant FWF Fund Science Austrian the thank to like would author The nti ae epoedorivsiaino h nepa between interplay the of investigation our proceed we paper this In oooia space topological A RDCSO EGRSAE NTEMILLER THE IN SPACES MENGER OF PRODUCTS S ahmtc ujc Classification. Subject Mathematics ia ocn oeswt uinadcmiaoilcvrn proper covering combinatorial and betwee fusion topology. interplay with s the posets answers of forcing instance This sical another spaces. gives and metrizable questions of open products finite by preserved Abstract. {∪V § ] oee,teaoestainbcmsipsil fw restrict we if impossible becomes situation above the However, 3]. n : n ∈ epoeta nteMle oe h egrpoet is property Menger the model Miller the in that prove We ω } or hspoet a nrdcdb ueiz n the and Hurewicz, by introduced was property This . σ cmatsaehsovosyteMne property, Menger the obviously has space -compact obntra oeigproperties covering combinatorial egrsae ocnrtdsae eile,Mle forcing. Miller semifilter, space, concentrated space, Menger hV X YBMRZDOMSKYY LYUBOMYR n a the has 1. : ,Y X, n Introduction X ∈ MODEL × ihacvrn rprymc stronger much property covering a with ω Menger Y i 1 rmr:0E5 42.Scnay 54C50, Secondary: 54D20. 03E35, Primary: uhthat such osnthv h Lindel¨of property, the have not does hU n : rpry(r lentvl,is alternatively, (or, property n X ∈ × V ,Y X, n Y ω ∈ i sntMne.There Menger. not is foe oesof covers open of [ U ⊂ n e,eg,[,17, [4, e.g., see, , ] R <ω hc aethe have which o all for ie nZFC In tive: clas- n isin ties everal cuniverse. ic vdi [15] in oved 1209-N25 I s covering l evdby served ountable aein rate e by red -called posets n r in ory tions have and X 2 LYUBOMYRZDOMSKYY are many results of this kind where CH is relaxed to an equality between cardinal characteristics, see, e.g., [3, 16, 23, 28]. Surprisingly, there are also inequalities between cardinal characteristics which imply that the Menger property is not productive even for sets of reals, see [27]. The following theorem, which is the main result of our paper, shows that an additional set-theoretic assumption in all these results was indeed necessary. Theorem 1.1. In the Miller model, the product of any two Menger spaces is Menger provided that it is Lindel¨of. In particular, in this model the product of any two Menger metrizable spaces is Menger. Theorem 1.1 answers [27, Problem 7.9(2)], [30, Problem 8.4] (restated as [31, Problem 4.11]), and [33, Problem 6.7] in the affirmative; implies that the affirmative answer to [1, Problem II.2.8], [5, Problem 3.9], and [16, Problem 2] (restated as [32, Problem 3.2] and [33, Problem 2.1]) is consistent; implies that the negative answer to and [1, Problem II.2.7], [6, Problem 8.9], and [36, Problems 1,2,3] is consistent; and answers [25, Problem 7] in the negative. By the Miller model we mean a generic extension of a ground model of GCH with respect to the iteration of length ω2 with countable support of the Miller forcing, see the next section for its definition. This model has been first considered by Miller in [20] and since then found numerous applications, see [11] and references therein. The Miller forcing is similar to the Laver one introduced in [18], the main difference being that the splitting is allowed to occur less often. The main technical part of the proof of Theorem 1.1 is Lemma 2.4 which is an analog of [18, Lemma 14]. The latter one was the key ingredient in the proof that all strong measure zero sets of reals are countable in the Laver model given in [18], the arguably most quotable combinatorial feature of the Laver model. As we shall see in Section 2, a big part of the proof of Theorem 1.1 requires only the inequality u < g which holds in the Miller model. However, we do not know the answer to the following Question 1.2. Is the Menger property preserved by finite products of metrizable spaces under u < g? If yes, can u < g be weakened to the Filter Dichotomy, NCF, or u < d? We refer the reader to [11, § 9] for corresponding definitions. We assume that the reader is familiar with the basics of forcing. The paper is essentially self-contained in the sense that we give all the definitions needed to understand our proofs.

2. Proofs First we collect some basic properties of Menger spaces. In most of the survey articles on combinatorial covering properties these are assumed to be a kind of folklore and are not stated (even without proofs) at all, see, e.g., [16, 30, 32, 33]. The second item of the following lemma is a partial case of [31, Proposition 2.1]. MENGER SPACES IN THE MILLER MODEL 3

Lemma 2.1. (1) If Z is a Menger space and X is a closed subspace of Z, then X is Menger when considered with the subspace topology inherited from Z. (2) If X is a countable family of Menger subspaces of a space Z, then S X is a Menger subspace of Z. (3) If X is Menger and K is compact, then X × K is Menger. Conse- quently, if X is Menger and Y is σ-compact, then X × Y is Menger. In particular, if X is Menger and Q is countable, then X × Q is Menger.

Proof. 1. Let hUn : n ∈ ωi be a sequence of open covers of X. For every U ∈ Un fix an open subset W (U) of Z such that W (U) ∩ X = U, and set Wn = {W (U): U ∈ Un}∪{Z \ X}. Applying the Menger property of Z to the sequence hWn : n ∈ ωi of open covers of Z we can find a sequence hVn : n ∈ ωi such that Vn is a finite subset of Wn and n∈ω Vn = Z. n n S S Write each Vn in the form {W (Ui ): itopological space X is called a

Gω1 -set if R is an intersection of ω1-many open subsets of X.

Definition 2.2. A topological space X is called weakly Gω1 -concentrated ω (resp. weakly ωGω1 -concentrated) if for every collection Q ⊂ [X] which 4 LYUBOMYRZDOMSKYY

is cofinal with respect to inclusion, and for every function R : Q → P(X) assigning to each Q ∈ Q a Gω1 -set R(Q) containing Q, there exists Q1 ∈ [Q]ω1 such that X ⊂ R(Q) (resp. for every Q ∈ [X]ω there exists SQ∈Q1 Q1 ∈ Q1 with the property Q ⊂ R(Q1)). Let A be a countable set and x, y ∈ ωA. As usually x ≤∗ y means that {a ∈ A : x(a) > y(a)} is finite. If x(a) ≤ y(a) for all a ∈ A, then we write x ≤ y. The smallest cardinality of a dominating with respect to ≤∗ subset of ωω is denoted by d. The smallset cardinality of a family B ⊂ [ω]ω generating an ultrafilter (i.e., such that {A : ∃B ∈ B (B ⊂ A)} is an ultrafilter) is denoted by u. By [7, Theorem 2] combined with the results of [10] the inequality ω1 = u < g = ω2 holds in the Miller model, see [7] or [11] for the definition of g as well as systematic treatment of cardinal characteristics of reals. As the following fact established in [21] shows, the inequality u < g imposes strong restrictions on the structure of Menger spaces. Lemma 2.3. In the Miller model, for every Menger space X ⊂ P(ω) and a Gδ-subset G such that X ⊂ G ⊂ P(ω), there exists a family K of compact subsets of G such that |K| = ω1 and X ⊂ S K. Consequently, in this model for every Menger space X ⊂ P(ω) and con- tinuous f : X → ωω there exists F ∈ [ωω]ω1 such that for every x ∈ X there exists f ∈ F with f(x) ≤ f. Proof. The first statement is [21, Theorem 4.4] combined with the fact that u = ω1 in the Miller model. Regarding the second statement, since the Menger property is preserved ω by continuous images and ω is homeomorphic to a Gδ-subset of P(ω), ω there exists a family K of compact subsets of ω such that |K| = ω1 and ω X ⊂ S K. For every K ∈K there exists fK ∈ ω such that y ≤ fK for all y ∈ K. It follows that the family F = {fK : K ∈ K} is as required.  By a Miller tree we understand a subtree T of ω<ω consisting of increas- ing finite sequences such that the following conditions are satisfied: • Every t ∈ T has an extension s ∈ T which is splitting in T , i.e., there are more than one immediate successors of s in T ; • If s is splitting in T , then it has infinitely many successors in T . The Miller forcing is the collection M of all Miller trees ordered by inclusion, i.e., smaller trees carry more information about the generic. This poset has been introduced in [20] and since then found numerous applications see, e.g., [10]. We denote by Pα an iteration of length α of the Miller forcing with countable support. If G is Pβ-generic and α < β, then we denote the intersection G ∩ Pα by Gα. For a Miller tree T we shall denote by Split(T ) the set of all splitting nodes of T , and for some t ∈ Split(T ) we denote the size of {s ∈ Split(T ): s ( t} by Lev(t, T ). For a node t in a Miller tree T we denote by Tt the set {s ∈ T : s is compatible with t}. It is clear that Tt is also a Miller tree. The stem of a Miller tree T is the (unique) t ∈ Split(T ) such that Lev(t) = 0. MENGER SPACES IN THE MILLER MODEL 5

We denote the stem of T by T h0i. If T1 ≤ T0 and T1h0i = T0h0i, then we 0 write T1 ≤ T0. The following lemma may be proved by an almost literal repetition of the proof of [18, Lemma 14].

Lemma 2.4. Let hx˙ i : i ∈ ωi be a sequence of Pω2 -names for reals and p ∈ ′ ′ 0 Pω2 . Then there exists p ≤ p such that p (0) ≤ p(0), and a finite set of reals ′ ′ Us for each s ∈ Split(p (0)), such that for each ε> 0, s ∈ Split(p (0)) with Lev(s,p′(0)) = i, j ≤ i, and for all but finitely many immediate successors t of s in p′(0) we have

′ ′ (p (0))t ˆp ↾ [1,ω2) ∃u ∈ Us (|x˙ j − u| <ε).

A subset C of ω2 is called an ω1-club if it is unbounded and for every α ∈ ω2 of cofinality ω1, if C ∩ α is cofinal in α then α ∈ C. Lemma 2.5. In the Miller model every Menger subspace of P(ω) is weakly 1 ωGω1 -concentrated (and hence also weakly Gω1 -concentrated ).

Proof. We work in V [Gω2 ], where Gω2 is Pω2 -generic. Let us fix a Menger space X ⊂ P(ω), consider a cofinal Q ⊂ [X]ω, and let R(Q) ⊃ Q be a

Gω1 -set for all Q ∈ Q. In the Miller model the Menger property is preserved by unions of ω1- many spaces, see [35, Theorem 4] and [7] for the fact that g = ω2 in this model. Combined with Lemma 2.1(1) this implies in particular that a com-

plement X \ R of an arbitrary Gω1 -subset R ⊂ P(ω) is Menger. Therefore by Lemma 2.3 and a standard closing off argument (see, e.g., the proof of [9, Lemma 5.10]) there exists an ω1-club C ⊂ ω2 such that for every α ∈ C the following condition is satisfied: ω Q∩V [Gα] is cofinal in [X] ∩V [Gα], and for every continuous ω f from a subset of P(ω) into ω such that f is coded in V [Gα], and every Q ∈ Q ∩ V [Gα] such that X \ R(Q) ⊂ dom(f), for ω every x ∈ X \ R(Q) there exists b ∈ ω ∩ V [Gα] such that f(x) < b.

Let us fix α ∈ C. We claim that Q ∩ V [Gα] has the required property. ˙ Suppose, contrary to our claim, there exists p ∈ Gω2 and a Pω2 -name Q∗ ω ω such that p forces “Q˙ ∗ ∈ [X˙ ] and Q˙ ∗ 6⊂ R˙ (Q˙ ) for any Q˙ ∈ [X˙ ] ∩ V [Γα]”, where Γα is the standard Pα-name for Pα-generic filter. There is no loss of generality2 in assuming that α = 0. Applying Lemma 2.4 to a sequence ′ ′ 0 hq˙i : i ∈ ωi enumerating Q˙ ∗, we get a condition p ≤ p such that p (0) ≤ ′ p(0), and a finite set Us of reals for every s ∈ Split(p (0)) such that for each ε > 0 and each s ∈ Split(p′(0)) with Lev(s,p′(0)) = i, for all but finitely

1 The proof of Theorem 1.1 requires only that Menger spaces are weakly Gω1 - concentrated in the Miller model. 2 This follows from the fact that the remainder P[α,ω2) of Pω2 evaluated by Gα, is forcing V [Gα] equivalent to Pω2 in V [Gα]. This can be proved in the same way as [18, Lemma 11]. 6 LYUBOMYRZDOMSKYY

many immediate successors t of s in p′(0) and all j ≤ i we have3 ′ ′ (1) p (0)t ˆp ↾ [1,ω2) ∃u ∈ Us (|q˙j − u| <ε). Note that any condition stronger than p′ will also satisfy condition (1), because for q0, q1 ∈ M the inequality q1 ≤ q0 implies Lev(s, q1) ≤ Lev(s, q0) Q ′ for all s ∈ Split(q1). Fix Q ∈ ∩ V containing X ∩ S{Us : s ∈ Split(p (0))} ′ and set F = X \ R(Q). It follows from the above that p Q˙ ∗ 6⊂ R˙ (Q). By passing to a stronger condition, if necessary, we may additionally assume ′ that p q˙j 6∈ R˙ (Q) for a given j ∈ ω. ′ Consider the map f : F → ωSplit(p (0)) defined as follows:

f(y)(s) = [1/ min{|y − u| : u ∈ Us}]+1 for4 all s ∈ Split(p′(0)) and y ∈ F . Since F is disjoint from Q, f is well- ′ defined. Since f is coded in V , there exist p′′ ≤ p′ and b ∈ ωSplit(p (0)) ∩ V ′′ such that p forces f˙(q ˙j) ≤ b. Without loss of generality we may additionally assume that Lev(p′′(0)h0i,p′(0)) ≥ j. Letting s′′ = p′′(0)h0i, we conclude ′′ ′′ ′′ that p f˙(q ˙j)(s ) ≤ b(s ), which means that ′′ ′′ p min{|q˙j − u| : u ∈ Us′′ } ≥ 1/b(s ). On the other hand, by our choice of p′, p′′ ≤ p′, condition (1), and Lev(s′′,p′(0)) ≥ j, for all but finitely many immediate successors t of s′′ in p′′(0) we have ′′ ′′ ′′ p (0)t ˆp ↾ [1,ω2) ∃u ∈ Us′′ |q˙j − u| < 1/b(s ) ′′ ′′ ′′ which means p (0)t ˆp ↾ [1,ω2) min{|q˙j − u| : u ∈ Us′′ } < 1/b(s ) and thus leads to a contradiction. 

The next lemma relates the weak Gω1 -concentration to products with Menger spaces. Lemma 2.6. In the Miller model, let Y ⊂ P(ω) be Menger and Q ⊂ P(ω) 2 be countable. Then for every Gω1 -subset O of P(ω) containing Q×Y there

exists a Gω1 -subset R ⊃ Q such that R × Y ⊂ O. Proof. Without loss of generality we shall assume that O is open. Let us write Q in the form {qn : n ∈ ω} and set On = {z ∈ P(ω): hqn, zi ∈ O} ⊃ Y . By Lemma 2.3 there exists a collection Z = {Zα : α ∈ ω1} of compact subsets of Tn∈ω On covering Y . It follows from the above that Q × Zα ⊂ O for all α. Applying [13, Lemma 3.1.15], for every q ∈ Q and α<ω1 we get an open Rq,α ⊂ P(ω) containing q such that Rq,α × Zα ⊂ O. Then Rα = Sq∈Q Rq,α is an open neighbourhood of Q in P(ω) with the property R × Z ⊂ O. Letting R = R we get that R × Y ⊂ R × Z ⊂ α α Tα<ω1 α Sα<ω1 α O. 

3Here we standardly identify P(ω) with the Cantor ternary subset of the unit interval, built by removing the middle thirds. This allows us to speak about |x − y| for elements x, y ∈P(ω). 4Here [a] is the largest integer not exceeding a. MENGER SPACES IN THE MILLER MODEL 7

As it was proved in [10], in the Miller model there exists an ultrafilter F generated by ω1-many sets, say {Fα : α ∈ ω1}. There is a natural linear ω pre-order ≤F on ω associated to F defined as follows: x ≤F y if and only if {n ∈ ω : x(n) ≤ y(n)}∈F. By [8, Theorem 3.1], in this model for every ω ω X ⊂ ω of size ω1 there exists b ∈ ω such that x ≤F b for all x ∈ X. n Lemma 2.7. In the Miller model, suppose that Un = {Uk : k ∈ ω} is an open cover of a Menger space X ⊂ P(ω), for every n ∈ ω. Then there exists b ∈ ωω such that X ⊂ U n for all α. (Equivalently, Sn∈Fα Sk≤b(n) k n {n ∈ ω : x ∈ Sk≤b(n) Uk }∈F for all x ∈ X.) Proof. The equivalence of two statements follows from the equality F = F +, where for X ⊂ [ω]ω we standardly denote by X + the set {Y ⊂ ω : Y ∩X =6 ∅ for all X ∈X}. To prove the second statement set G = Tn∈ω S Un and find a collection K of compact subsets of G such that |K| = ω1 and X ⊂ K ⊂ G. This S ω is possible by Lemma 2.3. For every K ∈ K find bK ∈ ω such that n ω K ⊂ U for all n ∈ ω. Then any b ∈ ω such that bK ≤F b for all Sk≤bK (n) k K ∈K is easily seen to be as required.  The second part of Theorem 1.1 is a direct consequence of Lemma 2.5 and the following Proposition 2.8. In the Miller model, let Y ⊂ P(ω) be a Menger space

and X ⊂ P(ω) be weakly Gω1 -concentrated. Then X × Y is Menger.

Proof. Fix a sequence hUn : n ∈ ωi of covers of X × Y by clopen subsets of 2 ω Q Q P(ω) . For every Q ∈ [X] fix a sequence hWn : n ∈ ωi such that Wn ∈ <ω Q [Un] and Q × Y ⊂ OQ,α for all α ∈ ω , where OQ,α = ∪W . (The 1 Sn∈Fα n latter is possible by Lemma 2.1(3) ensuring that Q×Y is Menger, combined with Lemma 2.7.) Letting O = O and using Lemma 2.6, we can Q Tα∈ω1 Q,α find a Gω1 -subset R(Q) ⊃ Q of P(ω) such that R(Q) × Y ⊂ OQ. Since ω ω1 X is weakly Gω1 -concentrated, there exists Q1 ∈ [[X] ] such that X ⊂ R(Q). For every Q ∈ Q let us find b ∈ ωω such that WQ ⊂{U n : SQ∈Q1 1 Q n k n k ≤ bQ(n)} for all n ∈ ω, where Un = {Uk : k ∈ ω} is an enumeration. Let ω b ∈ ω be an upper bound of {bQ : Q ∈ Q1} with respect to ≤F . We claim n Q that X × Y ⊂ Sn∈ω Sk≤b(n) Uk . Indeed, fix y ∈ Y , x ∈ X, and find Q ∈ 1 such that x ∈ R(Q). It follows that hx, yi ∈ OQ,α for all α ∈ ω1, therefore for Q n every α there exists n ∈ Fα with hx, yi ∈ ∪W ⊂ U , and hence n Sk≤bQ(n) k F := {n ∈ ω : hx, yi ∈ U n}∈F + = F. Then hx, yi ∈ U n Sk≤bQ(n) k Sk≤b(n) k for all n ∈ F ∩{k : bQ(k) ≤ b(k)}∈F, which completes our proof.  Note that Lemma 2.5 together with Proposition 2.8 imply that in the

Miller model, a subspace X of P(ω) is Menger iff it is weakly Gω1 -concen-

trated iff it is weakly ωGω1 -concentrated. As we have already noticed above, Lemma 2.5 and Proposition 2.8 imply Theorem 1.1 for subspaces of P(ω). The general case of arbitrary Menger spaces can be reduced to subspaces of P(ω) in the same way as in the proof of [24, Theorem 1.1], the only difference being that in some places 8 LYUBOMYRZDOMSKYY

“Hurewicz” should be replaced with “Menger”. However, we present this proof for the sake of completeness. We will need characterizations of the Menger property obtained in [35]. Let u = hUn : n ∈ ωi be a sequence of subsets of a set X. For every x ∈ X let Is(x,u,X) = {n ∈ ω : x ∈ Un}. If every Is(x,u,X) is infinite (the collection of all such sequences u will be denoted by Λs(X)), then we shall denote by Us(u,X) the smallest semifilter ω on ω containing all Is(x,u,X). (Recall that a family F ⊂ [ω] is called a semifilter if for every F ∈F and X∗ ⊃ F we have X ∈ F, where F ⊂∗ X means |F \ X| < ω.) By [35, Theorem 3], a Lindel¨of topological space X is Menger if and only if for every u ∈ Λs(X) consisting of open sets, the semifilter Us(u,X) is Menger when considered with the subspace topology inherited from P(ω). The proof given there also works if we consider only those hUn : n ∈ ωi ∈ Λs(X) such that all Un’s belong to a given base of X. Proof of Theorem 1.1. Suppose that X,Y are arbitrary Menger spaces such that X × Y is Lindel¨of and fix w = hUn × Vn : n ∈ ωi ∈ Λs(X × Y ) consisting of open sets. Set u = hUn : n ∈ ωi, v = hVn : n ∈ ωi, and note that u ∈ Λs(X) and v ∈ Λs(Y ). It is easy to see that

Us(w,X × Y )= {A ∩ B : A ∈ Us(u,X), B ∈ Us(v,Y )},

and hence Us(w,X × Y ) is a continuous image of Us(u,X) × Us(v,Y ). By [35, Theorem 3] both of latter ones are Menger, considered as subspaces of P(ω), and hence by Lemma 2.5 and Proposition 2.8 their product is Menger as well. Thus Us(w,X ×Y ) is Menger, being a continuous image of a Menger space. It suffices to use [35, Theorem 3] again. ✷ Acknowledgment. We would like to thank Arnold Miller for his comments regarding Lemma 2.4, and Marion Scheepers for the discussion of relations between Menger and productively Lindel¨of spaces. We thank the anony- mous referee for a careful reading and many suggestions which improved the presentation. References [1] Arkhangel’ski˘ı,A., Topological function spaces. Mathematics and its Appli- cations (Soviet Series), 78. Kluwer Academic Publishers Group, Dordrecht, 1992. [2] Aurichi, L.F., D-spaces, topological games, and selection principles, Topology Proc. 36 (2010), 107–122. [3] Babinkostova, L., On some questions about selective separability, Math. Log. Quart. 55 (2009), 539–541. [4] Bartoszynski, T.; Shelah, S.; Tsaban, B., Additivity properties of topological diagonalizations, J. Symbolic Logic 68 (2003), 1254–1260. [5] Bella, A., Bonanzinga, M., Matveev, M., Tkachuk, V., Selective separability: general facts and behaviour in countable spaces, Topology Proc. 32 (2008), 15–30. [6] Bernal-Santos, D.; Tamariz-Mascarua, A., The Menger property on Cp(X, 2), Topology Appl. 183 (2015), 110–126. [7] Blass, A.; Laflamme, C., Consistency results about filters and the number of inequivalent growth types, J. Symbolic Logic 54 (1989), 50–56. [8] Blass, A.; Mildenberger, H., On the cofinality of ultrapowers, J. Symbolic Logic 64 (1999), 727–736. MENGER SPACES IN THE MILLER MODEL 9

[9] Blass, A.; Shelah, S., There may be simple Pℵ1 - and Pℵ2 -points and the Rudin- Keisler ordering may be downward directed, Ann. Pure Appl. Logic 33 (1987), 213–243. [10] Blass, A.; Shelah, S., Near coherence of filters. III. A simplified consistency proof, Notre Dame J. Formal Logic 30 (1989), 530–538. [11] Blass, A., Combinatorial cardinal characteristics of the continuum, in: Hand- book of Set Theory (M. Foreman, A. Kanamori, and M. Magidor, eds.), Springer, 2010, pp. 395–491. [12] Chodounsk´y, D.; Repovˇs, D.; Zdomskyy, L., Mathias forcing and combinato- rial covering properties of filters, J. Symb. Log. 80 (2015), 1398–1410. [13] Engelking, R., General topology. Second edition. Sigma Series in Pure Math- ematics, 6. Heldermann Verlag, Berlin, 1989. [14] Gartside, P.; Medini, A.; Zdomskyy, L., The Tukey Order, Hyperspaces, and Selection Principles, work in progress. [15] Hurewicz, W., Uber¨ die Verallgemeinerung des Borellschen Theorems, Math. Z. 24 (1925), 401–421. [16] Just, W.; Miller, A.W.; Scheepers, M.; Szeptycki, P.J., The combinatorics of open covers. II, Topology Appl. 73 (1996), 241–266. [17] Koˇcinac, L.D.R., Selected results on selection principles. Proceedings of the 3rd Seminar on Geometry & Topology, 71–104, Azarb. Univ. Tarbiat Moallem, Tabriz, 2004. [18] Laver, R., On the consistency of Borel’s conjecture, Acta Math. 137 (1976), 151–169. [19] Menger, K., Einige Uberdeckungss¨atze¨ der Punktmengenlehre, Sitzungs- berichte. Abt. 2a, Mathematik, Astronomie, Physik, Meteorologie und Mechanik (Wiener Akademie) 133 (1924), 421–444. [20] Miller, A., Rational perfect set forcing, in: Axiomatic Set Theory (J. Baum- gartner, D. A. Martin, S. Shelah, eds.), Contemporary Mathematics 31, Amer- ican Mathematical Society, Providence, Rhode Island, 1984, pp.143–159. [21] Miller, A.W.; Tsaban, B.; Zdomskyy, L., Selective covering properties of prod- uct spaces, Ann. Pure Appl. Logic 165 (2014), 1034–1057. [22] Miller, A.W.; Tsaban, B.; Zdomskyy, L., Selective covering properties of prod- uct spaces, II: gamma spaces, Trans. Amer. Math. Soc. 368 (2016), 2865–2889. [23] Repovˇs, D.; Zdomskyy, L., On M-separability of countable spaces and function spaces, Topology Appl. 157 (2010), 2538–2541. [24] Repovˇs, D.; Zdomskyy, L., Products of Hurewicz spaces in the Laver model, Bull. Symb. Log. 23 (2017), 324–333. [25] Sakai, M.; Scheepers, M., The combinatorics of open covers. In: Recent progress in general topology. III (K. P. Hart, J. van Mill and P. Simon eds.), Atlantis Press, Paris, 2014, pp. 751–799. [26] Scheepers, M., Selection principles and covering properties in topology, Note Mat. 22 (2003/04), 3–41. [27] Szewczak, P.; Tsaban B., Products of Menger spaces: a combinatorial ap- proach, Ann. Pure Appl. Logic 168 (2017), 1–18. [28] Scheepers, M.; Tsaban, B., The combinatorics of Borel covers, Topology Appl. 121 (2002), 357–382. [29] Todorˇcevi´c, S., Aronszajn orderings. Duro Kurepa memorial volume, Publ. Inst. Math. (Beograd) (N.S.) 57(71) (1995), 29–46. [30] Tsaban, B., Selection principles in mathematics: A milestone of open prob- lems, Note Mat. 22 (2003/04), 179–208. [31] Tsaban, B., Additivity numbers of covering properties, in: Selection Principles and Covering Properties in Topology (L. Koˇcinac, eds.), Quaderni di Matem- atica 18, Seconda Universita di Napoli, Caserta 2006, 245–282. 10 LYUBOMYRZDOMSKYY

[32] Tsaban, B., Some new directions in infinite-combinatorial topology. In Set theory (J. Bagaria and S. Todorcevic, eds.), Trends Math., Birkh¨auser, Basel, 2006, 225–255. [33] Tsaban, B., Selection principles and special sets of reals, in: Open problems in topology II (edited By Elliott Pearl), Elsevier Sci. Publ., 2007, pp. 91–108. [34] Tsaban, B., Algebra, selections, and additive Ramsey theory, Fund. Math. 240 (2018), 81–104. [35] Zdomskyy, L, A semifilter approach to selection principles, Comment. Math. Univ. Carolin. 46 (2005), 525–539. [36] Zdomskyy, L, Can a Borel group be generated by a Hurewicz subspace? Mat. Stud. 25 (2006), 219–224.

Kurt Godel¨ Research Center for Mathematical Logic5, University of Vienna, Wahringer¨ Straße 25, A-1090 Wien, Austria. E-mail address: [email protected] URL: http://www.logic.univie.ac.at/~lzdomsky/

5Current address: Institute of Discrete Mathematics and Geometry, Technische Uni- versit¨at Wien, Wiedner Hauptstraße 8–10, 1040 Vienna, Austria